Movie Review Analysis based on Twitter

615 Final Project

GE JIN

December 16, 2017

Introduction

Introduction

My primary motivation was to explore reviews of different types of movies posted by users located in different regions based on the data source from Twitter. So I chose five recent movies, Murder on the Orient Express, Coco, Justice League, Daddy’s Home 2, Wonder, as my research subjects.

The analysis will help the moviegoers and the movie industry improve box-office results and help people easily decide what kind of movies to see basis their interests.

Exploratory Data Analysis

Number of Tweets

\(~\)

Movie Name Movie Type Release Date Number of Original Tweets Number of Retweets
Coco Animation 2017-11-22 35598 98873
Daddy’s Home 2 Comedy 2017-11-10 4828 4075
Justice League Action 2017-11-17 125830 164643
Murder on the Orient Express Mystery 2017-11-10 13556 4919
Wonder Drama 2017-11-17 7982 12302

Change of Number of Tweets over Time

Top Tweeters

Top Tweeters

Top Tweeters

Top Tweeters

Top Tweeters

Platforms of Users

Density of Original Tweets and Retweets

Retweet Network

Retweet Network of Orient Express

Retweet Network of Coco

Retweet Network of Justice League

Retweet Network of Daddy’s Home 2

Retweet Network of Wonder

Geolocations of Tweets

World Map

USA Map

States Map

States Map

States Map

States Map

States Map

Sentiment Analysis

Most Common Words about Orient Express

Word Sentiment Count
murder negative 16441
win positive 1224
mystery negative 481
love positive 236
death negative 215
amazing positive 201
top positive 186
enjoyed positive 175
pretty positive 169
classic positive 165

Most Common Words about Coco

Word Sentiment Count
win positive 21168
prize positive 14432
frozen negative 7821
perfect positive 3546
love positive 3045
beautiful positive 2923
critics negative 2742
cry negative 2236
amazing positive 2101
challenging negative 2047

Most Common Words about Justice League

Word Sentiment Count
win positive 7072
love positive 5804
marvel positive 5324
parody negative 4971
injustice negative 4308
effectively positive 3681
bad negative 3247
cool positive 3038
prize positive 2962
respect positive 2923

Most Common Words about Daddy’s Home 2

Word Sentiment Count
win positive 1518
prize positive 1450
funny negative 276
bad negative 205
hilarious positive 173
top positive 121
murder negative 114
fucking negative 61
love positive 57
recommend positive 56

Most Common Words about Wonder

Word Sentiment Count
win positive 496
magic positive 380
love positive 326
amazing positive 313
kindness positive 307
loved positive 282
beautiful positive 265
hug positive 251
wonderful positive 251
perfect positive 249

Sentiment Over Time about Orient Express

Sentiment Over Time about Coco

Sentiment Over Time about Justice League

Sentiment Over Time about Daddy’s Home 2

Sentiment Over Time about Wonder

Emojis Analysis

Emoji Reactions to Orient Express

Emoji Reactions to Coco

Emoji Reactions to Justice League

Emoji Reactions to Daddy’s Home 2

Emoji Reactions to Wonder

Conclusion

Conclusion

With the use of plethora of features in R, the analysis will provide key insights about those five movies, including timeline of tweets, user platforms of tweets, retweet network, etc. Also, maps were made to locate the tweets. The data was then reviewed graphically to explore the trending words in twitter about those movies and sentiment analysis was done using word cloud, line plots and horizontal bar charts, to tell us that how different customers react to different types of movies.

Various results and analysis showed that most people in USA prefer action movies, then animation movie, like Coco. The comedy Daddy’s Home 2 is least noticed in twitter. Besides, based on the states maps, people in California, Texas, Florida, New York and Virginia tweets more about those movies. But sentiment analysis, and emoji analysis, all together show us positive reactions of customers associated with the movies. win, love, prize, and Perfect - customers associate movie with positive words like these. This can help movie industry in understanding how branding and associating with these words can further help them improving box office.